S Aggarwal

Momentum for Smart City AI to make smarter cities, hospitals, infrastructures and companies

Discussion created by S Aggarwal on Oct 4, 2018

MomentumSelect Model

Select one or more from our ready-to-deploy AI model collection.
Deploy
Deploy Momentum AI on AWS, Google Cloud, IBM Cloud or docker containers on-premise.
Connect Data
Ingest and easily transform your data to make them ready for the AI.
Execute AI
Run AI either on-demand, in scheduled mode or in realtime on stream based data.
Visualize AI
Visualize using either our Momentum Insight, third party BI, or send to an existing app.

Momentum is a platform to rapidly deploy AI solutions. Momentum allows to ingest data from multiple sources and easily transform them to be able to feed to machine learning or AI algorithms to train models. The generated AI output or insights can be visualized using fully-integrated web based BI engine, Momentum Insight.

One single platform does it all -- from data ingestion to AI to visualization.

Momentum provides data ingestion from multiple sources and at scale.

Use SQL based interface to transform complex data -- both structured and unstructured - to quickly prepare your data for AI. This saves a significant amount of time so that you focus on model training to achieve a high level of accuracy.

Automate the data acquisition and transformation to take your experiment to production.

Momentum has implemented most valuable ML and AI algorithms so that AI developers simply bring data, select the algorithm, and execute to train the model -- all without writing a single line of code.

AI developers can also choose to train models in auto-pilot mode in which case Momentum selects the most appropriate models and train them using a set of hyper-parameter to find the optimum accuracy. The best model is selected for prediction.

AI model development is like performing experiments. After you get the perfect model, you would like to deploy that model in production. Momentum provides convenient ways to take your model from experiment to production.

For example, you can stream in realtime or schedule your data ingestion, transformation and prediction to run automatically and store the prediction output to an external system from where your existing application can display or make use of the AI output. Alternatively, you can display the insight in the form of dashboard using our fully-integrated visualization engine, Momentum Insight.

We expose a single Java Interface to extend the functionality of Momentum. A Java developer can implement the interface and write custom processing logic. Just drop the jar into Momentum and let Momentum takes care of distribution, parallelism, orchestration and overall job management.

Momentum is a cluster technology that scales as your data size and processing needs increase. Momentum can be deployed on cloud, such as Amazon AWS, Google Cloud, IBM Bluemix/Cloud, or Microsoft Azure. Momentum can also be deployed on premise in private cloud or data center.

or Smart City

AI to make smarter cities, hospitals, infrastructures and companies

Outcomes